MEMS, also known in the art as “micromachines,” are typically made up of individual components and generally range in size from 5 micrometers to a millimeter. They can consist of a central unit that processes data, the microprocessor and several components that interact with the outside such as microsensors.
According to prior art methods, the components of MEMS can be formed using photolithography and sacrificial layers. For example, multiple mask exposures which are capable of producing substantially arbitrary three-dimensional shapes are disclosed in U.S. Pat. No. 5,190,637 to Guckel (hereinafter “Guckel”), which is hereby incorporated by reference in its entirety. According to Guckel, a layer of photoresist capable of exposure by radiation is applied to a plating base. The photoresist is then exposed to radiation using a mask. The mask allows the radiation to only “expose” a certain defined area of the photoresist. Subsequent developing removes photoresist selective to the exposure creating a cavity that replicates the plan of the exposure mask. The cavity formed in the photoresist is then filled by a primary metal that is electroplated on to an exposed plating base. The remaining photoresist is then removed and a secondary metal (a sacrificial metal) is electroplated over the entirety of the primary metal and plating base. The primary metal and the secondary metal are then machined by mechanical means down to a height which exposes the primary metal and planarizes the surface for subsequent processing. After machining, another layer of photoresist can be applied across both the primary and secondary metals, and then this photoresist is also patterned using the same procedures as above. After the primary metal has been electroplated into the cavity created in the photoresist, the rest of the photoresist is removed and the secondary metal is electroplated over the entirety of the first secondary metal, any exposed first primary metal and the newly added second primary metal. Both the second primary and second secondary metals are machined down to the desired thickness of the second layer of the primary metal and the process is repeated until the desired number of layers have been formed creating the desired microstructure in the primary metal. Once the microstructure has been formed, the entirety of the plating surface together with the primary and secondary metals are exposed to an etching agent that selectively etches away the secondary metal but not the primary metal, thus leaving only the primary metal and the plating surface.
According to the teachings of Guckel, the secondary metal is used in conjunction with the photoresist because of the structural stability that it affords the primary metal during machining Machining cannot generally be done using only the photoresist and the primary metal because the photoresist is relatively weak mechanically and may not adequately support the primary metal from damage due to the largely lateral forces encountered in the mechanical machining process (which may include, grinding, lapping, polishing, chemo-mechanical polishing, electric discharge machining, or any other commonly encountered machining process). Another advantage the secondary metal affords is that it conveniently provides a conductive plating base for subsequent layers of primary metal that overhang underlying primary metal structures. Otherwise, the overhanging plating would require an additional thin film seed layer deposition step.
However, significant problems arise using the Guckel method when multiple or very laterally large microstructures are built on a single substrate, such as when manufacturing MEMS to be used as semiconductor testing probe heads. Plating the secondary metal across the entirety of the plating surface (e.g., ceramic) according to the teachings of Guckel causes the plating base to bow and warp under the stress of the additional sacrificial metal. This, in turn, causes two related problems: 1) it becomes difficult or impossible to machine the different layers to a uniform thickness and 2) it becomes difficult or impossible to perform the lithography because micro-lithography requires a planar surface.
U.S. Pat. No. 7,264,984 to Garabedian, et al. (hereinafter “Garabedian”) and U.S. Pat. No. 7,271,022, to Tang et al. (hereinafter “Tang”), both of which are incorporated by reference herein in their entireties, improved upon the teachings of Guckel by disclosing processes for creating micromechanical and MEMS structures, such that multiple or large structures may be built on the same substrate, without the substrate warping. More specifically, Tang discloses a method where the secondary metal is not plated over the entirety of the substrate's structured area. Instead the secondary metal is only plated in an area surrounding the primary metal structure, creating an “island,” so that it lends its structural stability to the primary metal structure, but does not cause undue stress on the substrate. According to the teachings in Tang, the sacrificial secondary metal is only plated where it is needed for mechanical stability and for the creation of structural overhangs, instead of requiring the secondary metal to be formed in the entire area encompassing all structural metal regions on the substrate.
Despite the improvements advanced in Garabedian and Tang, disadvantages remain. More specifically, the secondary metallization schemes of these references rely on photolithography to create a cavity immediately surrounding the primary metal and configured to hold the secondary metal. Photolithography is a time-consuming process that involves custom masks and photoresists to create the desired designs needed for plating the secondary metal. Accordingly, it is an objective of the teachings herein to provide a new method of MEMS fabrication that does not rely on photolithography with respect to plating secondary metals.
A further distinction between Garabedian and Tang and the present invention is that, neither of these references teach the use of secondary metals to cover the entire substrate for each layer, or anything less than a majority of layers, of a multi-layer intermediary MEMS structure. In contrast, both of these patents teach that if a secondary metal is plated over the entire surface of a substrate for too many levels, the substrate will warp. While these patents disclose, as non-preferred embodiments, the plating of secondary metal over the entire substrate for the first layer, or the initial layers of primary metal, they teach away from plating secondary metals over the entire substrate for each layer involved in a multi-layer MEMS fabrication process, in order to prevent warpage of the substrate. (See Tang, col. 7, lines 9-56). More explicitly, Tang stresses that the majority of secondary metal layers in a multi layered intermediary MEMS structure are not plated across the entire substrate. (See Tang, col. 7, lines 12-14)
Still another disadvantage of the methods provided in Garabedian and Tang is that their secondary metallization processes can lead to voiding in the secondary metal in between primary metal structures, using current deposition techniques. Voiding is undesirable as it minimizes the structural support the secondary metal provides the primary metal during the machining process. Due to the susceptibility of voiding in the secondary metallization process using current depositary techniques, there is a need in the art to provide a void-free secondary metallization scheme as part of the MEMS fabrication process.
While attempts have been made in the past to address voiding in the deposition of primary metals in cavities, such as during semi-conductor fabrication, these techniques focus on voiding in terms of electrical conductivity in metals that are integral to the final completed structure. Examples of superfilling primary metals for semi-conductor fabrication, are set forth in U.S. Pat. No. 6,946,716 to Andricacos, et al. (hereinafter “Andricacos”) and U.S. Pat. No. 6,432,821 to Dubin, et al. (hereinafter “Dubin”), both of which are incorporated herein in their entireties. The methodology provided in these disclosures is in sharp contrast to the present teachings, wherein superfilling is used with a secondary metal that surrounds the primary metal and is sacrificial, such that it is eventually etched away from the primary metal after the primary metal has been machined.
It is important to note that neither Andricacos nor Dubin address the problem of voiding in MEMS fabrication. More specifically, these references are silent as to the voiding of secondary sacrificial metals that surround and provide mechanical support to primary metals during machining. It is thus a further objective of the teachings herein to address this need in the art.